An overall goal of this three-year AREA grant application is to enhance the academic and research environment at Miami University by establishing a strong program in NMR- and MS-based metabonomics research for discovery of metabolic biomarkers for human diseases. The specific scientific goal of this research project will be to apply metabolic profiling to discover biomarkers for early detection of pancreatic cancer. More than 85% of lethal human pancreatic tumors originate in the ductal epithelia of the pancreas. Metabolic profiling will be applied to urine and fecal extracts collected from a mouse model for pancreatic cancer that recapitulates the full spectrum of pre-cancerous ductal neoplasias and tumor progression observed in humans. We predict that transformation of ductal epithelial cells during mPanIN initiation will generate a metabolic signature that will be detectable in urine and fecal extracts. Since it is currently not possible to detect human pancreatic cancer at the PanIN stage, we can only test this hypothesis and investigate the earliest metabolic signatures of PanIN initiation using this mouse model. Therefore, achieving the goals outlined in this proposal will only be possible using a mouse model that recapitulates the entire progression of PanIN initiation and tumor progression. In order that this study be most relevant to understanding initiation and progression of human pancreatic cancer, it is necessary to investigate a mouse model that closely reproduces the full spectrum of pre-invasive lesions and tumors that occur in human pancreatic cancer. The best mouse model available, which we will use in this study (Specific Aim #1) is the Pdx1-cre;LSL-KrasG12D transgenic strain, which has been well characterized and recapitulates the full range of pre- invasive pancreatic intraepithelial neoplasias observed in humans. The Pdx1-cre;LSL-KrasG12D strain develops a high incidence of mPanIN's by age 4-6 months and by the age of 6-12 months many of these mPanIN's transform to invasive pancreatic adenocarcinomas. The main goal of this project will be to conduct a 1-year-long longitudinal parallel metabonomics and MRI study to determine if a metabolic profiling biomarker can be discovered for early detection of pancreatic cancer (Specific Aim #4). The success of specific aim #4 will require accurate and reliable interpretation of MRI images of the mouse pancreas. To facilitate this goal, we have outlined a "road mapping" task to train to interpret MRI and MRJ images in the context of conventional histologies of the mouse pancreas (Specific Aim #3). Finally, given the complexity of the mouse pancreas, and the challenge presented by the fact that the majority of pancreatic tumors originate in the ductal epithelia, we have outlined a task to develop and demonstrate use of novel superparamagnetic antibody contrast agents targeted specifically to the pancreas ductal epithelia to maximize sensitivity and accuracy of detection in MRI images (Specific Aim #2). PUBLIC HEALTH RELEVANCE: Pancreatic cancer kills more than 35,000 individuals in the United States each year, is the fourth leading cause of cancer death in the USA, and is the most lethal form of cancer with about 70% of individuals dying within 4-6 months of initial diagnosis, and the five- year survival rate is less than 5%. The dismal outlook is because pancreatic cancer is usually not detected until it has spread and can no longer be treated by surgery. The best chance for improved survival of pancreatic cancer is early detection, but there are currently no effective biomarkers for early detection. The goal of this proposal is to use a novel technology of metabolic profiling to discover new biomarkers for early detection of pancreatic cancer, which if successful, could improve the survival and quality of life for thousands of Americans each year.